Misfolded GPI-Anchored Proteins Are Escorted Through the Secretory Pathway by ER-Derived Factors Eszter Zavodszky, Ramanujan S Hegde*
Total Page:16
File Type:pdf, Size:1020Kb
RESEARCH ARTICLE Misfolded GPI-anchored proteins are escorted through the secretory pathway by ER-derived factors Eszter Zavodszky, Ramanujan S Hegde* MRC Laboratory of Molecular Biology, Cambridge, United Kingdom Abstract We have used misfolded prion protein (PrP*) as a model to investigate how mammalian cells recognize and degrade misfolded GPI-anchored proteins. While most misfolded membrane proteins are degraded by proteasomes, misfolded GPI-anchored proteins are primarily degraded in lysosomes. Quantitative flow cytometry analysis showed that at least 85% of PrP* molecules transiently access the plasma membrane en route to lysosomes. Unexpectedly, time- resolved quantitative proteomics revealed a remarkably invariant PrP* interactome during its trafficking from the endoplasmic reticulum (ER) to lysosomes. Hence, PrP* arrives at the plasma membrane in complex with ER-derived chaperones and cargo receptors. These interaction partners were critical for rapid endocytosis because a GPI-anchored protein induced to misfold at the cell surface was not recognized effectively for degradation. Thus, resident ER factors have post-ER itineraries that not only shield misfolded GPI-anchored proteins during their trafficking, but also provide a quality control cue at the cell surface for endocytic routing to lysosomes. DOI: https://doi.org/10.7554/eLife.46740.001 Introduction Maintenance of a correctly folded proteome is critical for cellular and organismal homeostasis. Con- sequently, all cells employ protein quality control to identify and eliminate misfolded proteins (Wolff et al., 2014). The wide diversity of proteins and the multitude of compartments in eukaryotic *For correspondence: cells has driven the evolution of numerous quality control pathways for different classes of proteins [email protected] and different types of errors. Thus, an important step in understanding the principles of cellular pro- Competing interest: See tein homeostasis is to delineate the recognition and degradation pathways for major classes of mis- page 28 folded proteins. Funding: See page 28 Over 150 proteins in the human genome are anchored to the cell surface solely by a glycosyl- phosphatidylinositol (GPI) anchor in the membrane (UniProt Consortium, 2018). GPI-anchored pro- Received: 11 March 2019 Accepted: 15 May 2019 teins are ubiquitous across eukaryotes, often highly abundant, and have diverse roles including cell Published: 16 May 2019 adhesion, signaling, intercellular communication, and enzymatic reactions (Kinoshita et al., 2008). How misfolded GPI-anchored proteins are selectively recognized and degraded remains poorly Reviewing editor: Maya understood. The importance of this problem is highlighted by the capacity of mammalian prion pro- Schuldiner, Weizmann Institute, tein (PrP), a widely expressed and conserved GPI-anchored protein, to cause neurodegenerative dis- Israel ease when misfolded variants accumulate in cells (Prusiner, 2013). Copyright Zavodszky and The pathway used for degradation of a misfolded GPI-anchored protein depends on the step at Hegde. This article is distributed which its biosynthesis fails. Errors in targeting GPI-anchored proteins to the under the terms of the Creative endoplasmic reticulum (ER) or processing their signal for GPI anchor attachment at the membrane Commons Attribution License, which permits unrestricted use are handled by cytosolic (Ast et al., 2014; Hessa et al., 2011) and ER-associated degradation and redistribution provided that (ERAD) pathways (Ali et al., 2000; Ashok and Hegde, 2008; Sikorska et al., 2016; Wilbourn et al., the original author and source are 1998), respectively. Early studies in yeast suggested that once the GPI anchor is added, the mis- credited. folded protein is not degraded via Hrd1p (Fujita et al., 2006), a central ERAD factor that mediates Zavodszky and Hegde. eLife 2019;8:e46740. DOI: https://doi.org/10.7554/eLife.46740 1 of 30 Research article Cell Biology ubiquitination and retrotranslocation of substrates from the ER to the cytosol (Baldridge and Rapo- port, 2016; Bays et al., 2001; Stein et al., 2014). Instead, misfolded GPI-anchored proteins in yeast were suggested to use a seemingly unconven- tional pathway dependent on ER-to-Golgi transport (Fujita et al., 2006). Subsequent analysis sug- gested that ER export receptors of the TMED family (also known as the p24 family) rapidly sequester GPI-anchored proteins to prevent their engagement by Hrd1p, thereby allowing degradation in the vacuole (Sikorska et al., 2016). The GPI-anchored protein was primarily degraded by ERAD when the TMED cargo receptor was eliminated. Thus, the primary pathway for GPI-anchored protein deg- radation in yeast is via trafficking to the vacuole, with ERAD serving as a failsafe pathway when the vacuole route is impaired. Parallel studies in mammalian cells investigating mutant PrP degradation arrived at similar conclu- sions. First, investigation of the localization, trafficking, and turnover of a panel of human disease- causing PrP mutants showed that they are not degraded by ERAD, do not depend on the protea- some, and exit the ER despite their misfolding (Ashok and Hegde, 2009). The misfolded population of mutant PrP was selectively observed in post-ER intracellular compartments en route to their ulti- mate degradation in acidic compartments presumed to be lysosomes. Using an artificial constitu- tively misfolded PrP mutant (termed PrP*, containing a C179A mutation that cannot form the sole disulfide bond in PrP), trafficking from the ER to lysosomes was directly visualized by time-lapse imaging in live cells (Satpute-Krishnan et al., 2014). This study showed that PrP* is primarily retained in the ER at steady state but can be released into the secretory pathway by acute ER stress. The steps between ER retention and lysosomal clearance are only partially understood. Transit of PrP* to the Golgi requires cargo receptor TMED10 (also known as Tmp21, or p24d1) with which it interacts in co-immunoprecipitation experiments (Satpute-Krishnan et al., 2014). From here, the route to lysosomes is not established. At least a subpopulation was implicated in transiting the cell surface based on extracellular antibody uptake assays and trapping of PrP* at the cell surface after cholesterol depletion (Satpute-Krishnan et al., 2014). The proportion of PrP* using this itinerary was unclear but it is important to understand because exposing misfolded proteins to the extracellu- lar environment can be detrimental. In the specific case of PrP, surface-exposed misfolded forms may facilitate uptake of prions into cells (Fehlinger et al., 2017). From these combined studies in yeast and mammalian cells, it is thought that both folded and misfolded GPI-anchored proteins engage TMED family export receptors at the ER and traffic to the Golgi. At some step at or after the trans-Golgi network, their itineraries diverge. Folded GPI- anchored proteins go on to reside at the cell surface, whereas misfolded variants are delivered to the lysosome. It is not known how misfolded GPI-anchored proteins get from the Golgi to lyso- somes, how they avoid aggregation during their journey through chaperone-poor post-ER compart- ments, or how cells discriminate folded from misfolded proteins to influence their trafficking. Here, we used quantitative flow cytometry and proteomic analyses to show that the majority of PrP* traf- fics via the cell surface to lysosomes in a complex with resident ER chaperones and cargo receptors. This suggests that minor populations of abundant factors long thought to be restricted to the early secretory pathway have functional excursions to the cell surface during quality control of GPI- anchored proteins. Results Experimental system for quantitative analysis of PrP* degradation To perform quantitative analysis of misfolded GPI-anchored protein degradation, we first generated and characterized a stable doxycycline-inducible HEK293T cell line expressing GFP-tagged PrP* (GFP-PrP*) integrated into a single defined locus in the genome. This mutant of PrP contains a Cys to Ala change at position 179, thereby preventing the formation of a critical disulfide bond required for PrP folding (Satpute-Krishnan et al., 2014). A matched cell line expressing wild type GFP-PrP from the same locus served as a control in these studies. Immunoblotting of total cell lysates after induction with doxycycline showed that the steady state level of GFP-PrP* was very similar to GFP- PrP (Figure 1A). The different migration patterns are due to complex glycosylation of GFP-PrP dur- ing its transit through the Golgi in contrast to core-glycosylated GFP-PrP* primarily retained in the ER. Zavodszky and Hegde. eLife 2019;8:e46740. DOI: https://doi.org/10.7554/eLife.46740 2 of 30 Research article Cell Biology A GFP- GFP- CD PrP PrP* dox: - + - + GFP-PrP* DMSO 75 complex Tg GFP 58 core Tg + Baf (faint) 46 no glyc. GFP 75 complex 58 core (dark) 46 no glyc. total proportion of cells protein GFP-PrP* -10 3 010 3 10 4 10 5 B uninduced GFP-PrP GFP-PrP* GFP-PrP DMSO Tg Tg + Baf proportion of cells proportion of cells GFP-PrP 3 3 4 5 -103 010 3 10 4 10 5 -10 010 10 10 GFP fluorescence GFP fluorescence E 0 min 30 min 60 min 120 min 180 min GFP-PrP* Figure 1. A stable-inducible cell line to study GPI-anchored protein quality control. (A) HEK293-TRex cells containing either GFP-PrP or GFP-PrP* stably integrated at the same locus were induced to express the proteins with doxycycline for 48 hr prior to analysis by immunoblotting using anti-GFP antibody. Cultures without doxycycline induction were analyzed in parallel. Two exposures of the immunoblot are shown, along with a portion of the stained blot verifying equal loading. Unglycosylated (‘no glyc.”), core-glycosylated (‘core’), and complex-glycosylated (‘complex’) species of GFP-PrP are indicated. (B) Cells expressing GFP-PrP or GFP-PrP* were induced with doxycycline for 96 hr prior to analysis of GFP fluorescence by flow cytometry. The normalized histograms are shown. (C) Cells expressing GFP-PrP or GFP-PrP* were induced for 48 hr with doxycycline, washed to remove doxycycline, then analyzed 24 hr later by fluorescent microscopy.